DESCRIPTION: (Applicant's Abstract) Cellular injuries caused by cytotoxic agents produce a highly regulated "cellular injury response". One characteristics of this response is the increase in mRNA levels of many genes. Using this characteristic the applicant demonstrated that a cisplatin-induced increase in the mRNA level of one of the genes involved in the "cellular injury response", GADD153, correlated well with tumor cell kill in vitro and with response rates in individual patients. Instead of examining the relationship between a particular phenotype and changes in the level of one mRNA at a time, new techniques now allow for the identification of many more genes whose mRNA levels change during the cellular injury response. This approach is based on the use of suppressive subtractive hybridization to isolate differentially regulated messages combined with high throughput screening of cDNA microarrays and will allow the generation of an expression pattern (fingerprint) by measuring the levels of up to several thousand mRNAs at the same time. This project will focus on the description of fingerprints caused by a specific drug, cisplatin, in a specific tumor type, squamous cell carcinomas, to validate a high throughput screening approach measuring the changes of many mRNA levels during the response of tumor cells to injury. Therefore, it is the applicant's hypothesis that specific changes in mRNA levels of genes responsive to injury can be recognized as "fingerprints" and that these fingerprints will include essential information about the responsiveness of tumor cells to treatment. The Specific Aims are to: 1) to identify genes whose mRNAs undergo changes in abundance measured at 4, 8, 24, 48, and 72 hours following exposure of head and neck cancer cells to cDDP at dose levels sufficient to result in 10% survival using the technique of Suppression Subtractive Hybridization (SSH). This will generate five libraries of cDNA fragments derived from differentially expressed messages following injury. 2) Identify those elements which follow a similar expression pattern after hybridization of each of the five libraries with probes generated after 2,4,8 12, 24,48, 72, and 96 hours following a one hour exposure of cDDP. 3) Determine whether the cDDP- induced injury in head and neck cancer cells can be identified on the basis of an expression pattern (fingerprint). Replicated copies of the array containing truly differentially expressed fragments and grouped accordingly to their expression profile (Specific Aim 2) will be probed with labeled cDNA generated from the original head and neck cancer cell line (UMSCC10b), and from 10 additional head and neck cancer cell lines. The fingerprints will be compared to identify a subset of cDNAs whose differential expression reliably detects the injury response to cDDP in vitro. 4) To determine whether expression signatures that identify the cDDP-injury phenotype in vitro are capable of identifying this phenotype when the cells are grown in vivo in nude mice. The results of this project are expected to provide the foundation for the development of a clinically applicable technology to determine tumor response based on changes in expression patterns of RNA levels. In addition, an important byproduct of this research will be the identification of genes whose expression is causatively linked to pathways leading to cell death and which are thus potential new pharmaceutical targets.